Selective DOT1L, LSD1, and HDAC class I inhibitors reduce HOXA9

Subscriber access provided by UNIV OF DURHAM

Article

Selective DOT1L, LSD1, and HDAC class I inhibitors reduce HOXA9 expression in MLL-AF9 rearranged leukemia cells, but dysregulate the expression of many histone-modifying enzymes Ryan Lillico, Courtney K Lawrence, and Ted M. Lakowski J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00118 • Publication Date (Web): 04 Jul 2018 Downloaded from http://pubs.acs.org on July 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

1 Selective DOT1L, LSD1, and HDAC class I inhibitors reduce HOXA9 expression in MLL-AF9 rearranged leukemia cells, but dysregulate the expression of many histone-modifying enzymes

Ryan Lillico1, Courtney K. Lawrence1 and Ted M. Lakowski1*. 1

Rady Faculty of Health Science, College of Pharmacy, Pharmaceutical Analysis Laboratory, University

of Manitoba, 750 McDermot Avenue, Winnipeg, Manitoba, Canada, R3E 0T5.

Abstract Mixed lineage leukemia results from chromosomal rearrangements of the gene mixed lineage leukemia (MLL). MLL-AF9 is one such rearrangement that recruits the lysine methyltransferase, human disruptor of telomere silencing 1-like (DOT1L) and lysine specific demethylase 1 (LSD1), resulting in elevated expression of the Homeobox protein A9 (HOXA9), and leukemia. Inhibitors of LSD1 or DOT1L reduce HOXA9 expression, kill MLL-rearranged cells, and may treat leukemia. To quantify their effects on histone modifying enzyme activity and expression in MLL-rearranged leukemia, we tested inhibitors of DOT1L (EPZ-5676), LSD1 (GSK2879552), and HDAC (mocetinostat), in the MLL-AF9 cell line MOLM-13. All inhibitors reduced MOLM-13 viability but only mocetinostat induced apoptosis. EPZ5676 increased total histone lysine di-methylation, which was attributed to a reduction in LSD1 expression, and was indistinguishable from direct LSD1 inhibition by GSK2879552. All compounds directly inhibit, or reduce the expression of, HOXA9, DOT1L and LSD1 by qPCR, increase total histone lysine methylation and acetylation by LC-MS/MS, and specifically reduce H3K79Me2 and increase H3K14Ac. Each inhibitor altered the expression of many histone modifying enzymes which may precipitate additional changes in expression. To the extent that this decreases HOXA9 expression it benefits mixed lineage leukemia treatment, all other expression changes are off-target effects. Keywords: Epigenetics, MLL-rearranged leukemia, histone modifications, LC-MS/MS

ACS Paragon Plus Environment

Journal of Proteome Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

2 1. Introduction The disruptor of telomere silencing 1-like (DOT1L), a histone H3 lysine 79 (H3K79) methyltransferase, is a regulator of gene transcription, somatic reprogramming, DNA damage response, and cell cycle regulation.1, 2 DOT1L is the only H3K79 methyltransferase known. It contains a DNA binding domain and only methylates nucleosomal substrates and not free histones,3 suggesting that H3K79 methylation is involved in chromatin remodeling rather than nucleosome assembly. H3K79 dimethylation fluctuates through cell-cycle checkpoints and is found mostly in actively transcribed genes, implicating DOT1L as a regulator of euchromatin, transcriptional elongation, and cellular differentiation.4 DOT1L inhibitors are being investigated as treatments for leukemia that result from rearrangements of the gene, mixed lineage leukemia (MLL).5 MLL is a H3K4 methyltransferase and its wild-type activities control expression of the homeobox A cluster (especially HOXA9), which are involved in normal hematopoiesis.6 Translocations and insertions at chromosome 11q23 results in fusions of MLL with a variety of partners (among these are MLL-AF9, MLL-AF10, and MLL-AF4) that eliminate its H3K4 methylation activity but retain its target gene DNA binding capacity.7 The translocation t(9;11)(p22;q23) or insertion ins(11;9)(q23;p22p23) result in the fusion of MLL to AF9, (MLL-AF9). MLL-AF9 recruits DOT1L to the MLL DNA binding sites resulting in the replacement of H3K4 with H3K79 methylation within the HOXA9 gene and promoter.8 This sustains the overexpression of HOXA9 resulting in the inability to differentiate which produces a self-renewing cell population and mixed lineage leukemia.6 DOT1L inhibitors have been shown to decrease HOXA9 expression and increase cellular differentiation, resulting in the selective killing of MLL-rearranged cells. The DOT1L inhibitor EPZ-5676 is under investigation as a treatment for MLL-rearranged leukemia.9 However, DOT1L is an important regulator of the expression of multiple genes, therefore even its selective inhibition may lead to unpredictable changes in the expression of multiple genes. We hypothesize that such changes in expression may include other histone modifying enzymes, which would manifest itself as changes in total histone and residue specific histone modifications and this could in turn precipitate down-stream changes in gene expression. It has been shown previously that reducing HOXA9 expression with shRNA knockdown is sufficient on


Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3 its own to arrest the growth of MLL-AF9 cell lines.10 Therefore, changes in the expression of genes that don’t ultimately reduce HOXA9 expression, are by definition off-target effects. We and others have observed that histone deacetylase (HDAC) inhibitors, alter the expression of several genes.11 It is therefore possible that drugs targeting other histone modifying enzymes may indirectly decrease HOXA9 expression and may be useful as treatments for mixed lineage leukemia as long as other induced off-target changes in histone modifying expression do not result in toxicity. The lysine specific demethylase 1 (LSD1 or KDM1A) acts on H3K4 and H3K9 and can only demethylate mono- and dimethyl-lysine because of its FAD dependent mechanism.12 LSD1 is a partner in co-repressor complexes (CoREST, NuRD, CtBP), co-activator super complexes (MLL, ELL)13 and is important in organogenesis14 and hematopoiesis.15 LSD1 is thought to be involved with the MLL-AF9 complex, supporting the activation of HOXA9 thereby preventing differentiation, and as a result increases the leukemic cell population.16 Inhibiting LSD1 promotes hematopoietic differentiation,17 resulting in toxicity to MLL-rearranged cells, much like DOT1L inhibition. Interestingly, administering LSD1 and DOT1L inhibitors together, synergistically inhibits proliferation of MLL-rearranged leukemia cells.17 DOT1L and LSD1 are cell cycle and proliferation regulators. Inhibition of DOT1L leads to G0/G1 arrest18, 19

and accumulation of the cell population in the sub-G1 phase,9 depletion of LSD1 partially stalls cells at

the G2/M checkpoint20 and knockout of either LSD1 or DOT1L in pluripotent cells induces death.2, 21 Previously, we have shown that HDAC inhibitors can reduce the expression of LSD1 in addition to other changes in gene expression.11 This suggests that HDAC inhibitors may be used for similar indications as LSD1 inhibitors. In this study, we tested inhibitors of histone modifying enzymes in an acute myeloid leukemia cell line containing the MLL-AF9 fusion, that overexpresses HOXA9 (MOLM-13), compared to a chronic myeloid leukemia cell line (BCR-ABL1 fusion) that has low HOXA9 expression (K562). We evaluated the effects of representative selective HDAC class I (mocetinostat), DOT1L (EPZ-5676), and LSD1 (GSK2879552) inhibitors on LSD1 and HOXA9 expression, total and residue specific histone posttranslational modifications, cell viability, apoptosis, and the expression of an array of histone modifying



Page 4 of 31

4 enzymes. EPZ-5676 reduced expression of HOXA9 in MOLM-13 cells and reduced viability when treated for 6 days as shown previously.9 Despite the fact that it is a lysine methyltransferase inhibitor, EPZ-5676 increased total histone lysine di- and monomethylation, similar to the LSD1 inhibitor GSK2879552. In an effort to explain this activity, we found that EPZ-5676 reduced LSD1 expression. We also confirmed that GSK2879552 decreased HOXA9 expression. EPZ-5676 and GSK2879552 decreased the viability of MOLM-13 cells but had no effect on K562 cells. We have shown previously that the selective HDAC class I inhibitor mocetinostat decreases LSD1 expression in K562 cells. To see if this could produce effects similar to the direct LSD1 inhibition produced by GSK2879552 and the reduction in LSD1 expression produced by EPZ-5676, we treated MOLM-13 cells with mocetinostat finding that it decreased LSD1 and HOXA9 expression and induced apoptosis in both cell lines. EPZ-5676 is the most effective inhibitor studied at decreasing HOXA9 expression in MOLM-13 cells. Despite the difference in histone modifying enzyme targets, all three compounds directly inhibit, or reduce the expression of HOXA9, DOT1L and LSD1, increase total histone lysine methylation and acetylation and specifically reduce H3K79Me2 and increase H3K14Ac. As predicted, the three compounds changed the expression of several histone-modifying enzymes, which themselves are involved in controlling gene expression. Where this results in reduced HOXA9 expression these changes in expression are advantageous for the treatment of mixed lineage leukemia. However, where such activities do not eventually decrease HOXA9 expression they can be viewed as off target effects. 2. Materials and Methods 2.1 Cell culture and treatments MOLM-13 cells carrying MLL-AF9 fusion from ins(11;9)(q23;p22p23) (DSMZ ACC 554 ref number is A1408102-1) were cultured in RPMI 1640 media supplemented with 1% penicillin/streptomycin and 10% FBS. K562 cells (ATCC (CCL-243)) were cultured in IMDM media with 1% penicillin/streptomycin and 10% FBS. Cells were incubated at 37 °C in 5% CO2 and were passaged every 72 h. Cells (5x104/mL) were treated with various concentrations of the DOT1L inhibitor, EPZ-5676, the LSD1 inhibitor, GSK2879552, or the HDAC inhibitor mocetinostat (stocks in 100% DMSO, final concentration in media


Page 5 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5 0.1% DMSO) for 24, 72, or 144 h. For 144 h incubations, cells were counted and diluted back to 5x104/mL in fresh media containing the appropriate concentration of inhibitor after 72 h. MOLM-13 cells derived from the blood cells of a 20-year old male with mixed lineage leukemia were used as our model cell line because they harbor the MLL-AF9 fusion and overexpress HOXA9. K562 cells were used as a control cell line because they have undetectable expression of HOXA9. (For additional information on cell lines and for cell authentication see Supplemental material). 2.2 Cell viability and apoptosis MOLM-13 and K562 cells were plated 5x104/well in 96-well format and treated with EPZ-5676 for 144 h, GSK2879552 for 72 h, and mocetinostat for 24 h. Cell viability and apoptosis were then evaluated using the ApoTox-Glo (Promega) assay according to manufacturers instructions. Fluorescence (viability) and luminescence (apoptosis) were measured using the BMG FLUOstar Galaxy microplate reader. 2.3 Histone modification analysis Cells were harvested, and histones isolated using the Epiquik (Epigentek) kit with modifications. Histones were precipitated with 4% perchloric acid at 4 °C overnight, washed with acetone and reconstituted in water to 1 mg/mL. Isolated histone samples (10 µg) were digested using Pronase or hydrolyzed with 6N HCl vapor and analyzed for acetyl-lysine, methyl-lysines and methyl-arginines by liquid chromatography tandem mass spectrometry (LC-MS/MS) according to previously established protocols.11 Briefly, the modified and unmodified amino acids from the histone hydrolysates were separated on a Primesep 200 (Sielc) column with a pH gradient in acetonitrile and water using a Shimadzu Nexera UHPLC. Each analyte was measured in positive MRM mode using a Shimadzu LCMS 8040 mass spectrometer with DUIS ionization and quantified using synthetic standards of acetyl-lysine, mono-, di-, and trimethyllysine, monomethyl-arginine, asymmetric dimethyl-arginine and symmetric dimethyl-arginine. Each sample was normalized to the quantity of lysine and each modification was expressed as percent change with respect to controls receiving no treatment. Specific modifications on histone H3 including H3K4Me1-3, H3K9Me1-3, H3K27Me1-3, H3K36Me1-3, H3K79Me1-3, H3K9Ac, H3K14Ac, H3K18Ac, H3K56Ac, H3Ser10P and H3Ser28P were measured



Page 6 of 31

6 using the Histone H3 Modification Multiplex Assay Kit (Abcam, ab185910) according to the manufacturer’s instructions, using 50 ng histone isolate per well. Each Treatment was normalized to total H3 and the changes in modifications were calculated as a percentage with respect to a no treatment control. 2.4 qPCR gene expression analysis Total mRNA was isolated and purified from cells using the Ambion Pure Link RNA isolation kit (Life Technologies) and cDNA synthesized using Super Script Vilo master mix (life Technologies). Quantitative PCR (qPCR) was performed using the Applied Biosystems Viia7 real-time PCR system with TaqMan assays (Life Technologies, Supplemental Table 1). PCR reactions were prepared in 10 or 20 µL volumes using 50-100 ng cDNA in TaqMan fast universal master mix (Life Technologies) following the manufacturers conditions. Relative gene expression (RQ) was quantified by comparative CT (∆∆CT) to no treatment controls and normalized using either Beta-2Microglobulin (B2M) or 18 s rRNA housekeeping genes. Dose versus gene expression (response) curves were fit to a 4 parameter logistic regression using SigmaPlot 11 according to the guidelines of the NIH National Center for Advanced Translational Sciences and the IC50, min and max determined.22 2.5 In vitro DOT1L methylation assay and LSD1 inhibitor-screening assay Increasing concentrations of EPZ-5676 were incubated for 1 hour with 120 nM recombinant DOT1L (EpiCypher), 1 µg of recombinant nucleosomes (EpiCypher) and 24 µM S-adenosyl methionine (AdoMet) in 100 µL of a methylation buffer containing 50 mM ammonium acetate, 2.5 mM NaCl, 0.01% Triton X100 and 0.25 mM DTT. The reactions were dried in a vacuum centrifuge and the products hydrolyzed with 6 N HCl vapor for 24 hours and analyzed for methyl-lysines by LC-MS/MS according to previous methods.11 DOT1L activity was represented as the weighted sum of all methyl groups transferred to lysine. Measurements of the drug treatments were compared to a DMSO control (100% activity) and no AdoMet control (0% activity).


Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7 EPZ-5676, mocetinostat and GSK2879552 were screened for LSD1 inhibition using the LSD1 inhibitorscreening assay (Cayman, Cat. 700120) according to the manufacturer’s instructions. Absorbance was measured using the BioTek Synergy HT microplate reader. 3. Data accessibility The ChIP-seq data sets used to make Supplemental Figure S2 were originally published23 and graphics of these data were generated using the Integrative Genomics Viewer 2.3.77 using tracks loaded from the Encyclopedia of DNA elements (ENCODE) (K562 cells) along with the available data sets from the NCBI

Gene

expression

omnibus

(GEO)

database

for

MOLM-13

cells

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE76750. 4. Results 4.1 Inhibition of DOT1L or LSD1 reduces HOXA9 expression in MOLM-13 cells Previous studies have shown that reversing overexpression of HOXA9 with shRNA in cell lines harboring the MLL-AF9 translocation, is alone sufficient to force differentiation and arrest growth, eventually leading to cell death.10 Therefore, reducing HOXA9 expression is a validated target for treatment of MLL-rearranged leukemia and the effectiveness of potential new treatments can, in part be evaluated by measuring reductions in HOXA9 expression. MOLM-13 cells were chosen as a model for mixed lineage leukemia with the MLL-AF9 fusion. Initially the expression of HOXA9 was evaluated using qPCR in MOLM-13 cells that overexpress HOXA9 and compared to K562 cells that express undetectable levels of HOXA9,24 with and without treatment with EPZ-5676, in order to validate the qPCR method (supplemental Figure S1). MOLM-13 cells were treated with mocetinostat, EPZ-5676, and GSK2879552 and we found that all three reduced HOXA9 expression in a dose dependent fashion (Figure 1A, Table1). Similar to previous studies with EPZ-5676,9 the maximum reduction in HOXA9 expression (>90%) occurred at longer incubation periods up to 144 h with 5 µM, while the same dose at 24 h and 72 h reduced HOXA9 expression approximately by 30% and 80%, respectively (Figure 1B). The IC50 for the reduction in HOXA9



Page 8 of 31

8 expression with a 72 h EPZ-5676 treatment was 150±10 nM, which is within the same range of previously calculated values in MLL-AF4 rearranged MV4-11 cells (Table 1).9 The effects of the LSD1 inhibitor GSK2879552 were evaluated in MOLM-13 cells because such tranylcypromine analogue inhibitors have, (like DOT1L inhibitors) recently been identified as potential treatments for MLL-rearranged leukemia that reduce HOXA9 expression.16 GSK2879552 reduced HOXA9 expression evaluated with increasing treatment time and dose producing a maximum ~40% reduction in HOXA9 expression at a 72 h incubation period (Figure 1B). Longer incubation periods did not result in greater reduction of HOXA9 expression as seen with the DOT1L inhibitor EPZ-5676. Although the maximum decrease in HOXA9 expression by GSK2879552 treatment only resulted in approximately 40% reduction, it did so potently with an IC50 of 10±3 nM at 72 h (Figure 1A, Table 1). Having confirmed that the LSD1 inhibitor GSK2879552 reduced HOXA9 expression we evaluated whether the Class I HDAC inhibitor mocetinostat could reduce HOXA9 expression because we have previously shown that it reduces the expression of several lysine demethylases in K562 cells, and in particular LSD1.11 Mocetinostat rapidly reduced HOXA9 expression by up to 50% in 24 h in MOLM-13 cells with a similar potency to EPZ-5676 (IC50 of 105±53 nM) (Figure 1A, Table 1). As far as we know this is the first time that any HDAC inhibitor has been shown to reduce HOXA9 expression in an MLLAF9 cell line. 4.2 EPZ-5676 and GSK2879552 are toxic to MOLM-13 but not K562 cells Viability and apoptosis for the selected inhibitors in both MOLM-13 and K562 cells were measured using the ApoTox-Glo assay. EPZ-5676 and GSK2879552 reduced the viability of MOLM-13 cells, but not K562 (Figure 2A). A 144 hour treatment of EPZ-5676 and a 72 hour treatment of GSK2879552 did not induce caspase activation in either cell type, suggesting the reductions in viability were through nonapoptotic mechanisms. However, mocetinostat induced apoptosis in both cell types in 24 hours (Figure 2B). All three inhibitors showed a dose dependent reduction in viability in MOLM-13 cells, but only mocetinostat did so in K562 cells (Figure 2C and D). Despite not inducing complete reductions in


Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


9 HOXA9 expression, mocetinostat appeared to be more potent and ultimately more effective than EZP5676 with respect to reductions in MOLM-13 cell viability (Figure 2C and D and Table 1). 4.3 Mocetinostat, EPZ-5676, and GSK2879552 produce similar changes in total histone modifications MOLM-13 cells were treated with the HDAC class I inhibitor mocetinostat, the DOT1L inhibitor EPZ5676, and the LSD1 inhibitor GSK2879552 to determine if the inhibitors produced changes to total histone modifications (Figure 3A). Total histone modifications were measured at the time and dose that each inhibitor yields its maximum HOXA9 inhibition (100 nM, 24 h for mocetinostat, 3 µM, 144 h for EPZ5676 and 1 µM, 72 h for GSK2879552) using our previously developed LC-MS/MS assay.11 Surprisingly, we observed similar trends in total histone modifications for all three inhibitors. Given that EPZ-5676 is a methyltransferase inhibitor we expected to see reductions in histone lysine methylation but paradoxically, increased lysine dimethylation, and to a lesser extent monomethylation, were observed (Figure 3A). Such changes in histone lysine methylation could indicate increasing activity or expression of methyltransferases catalyzing mono- and di-methylation, or decreases in the activity or expression of FAD dependent demethylases such as LSD1 since these enzymes are only capable of demethylating mono and dimethyl-lysine.12 Accordingly, the potent LSD1 inhibitor GSK2879552, resulted in a similar pattern of increasing total histone lysine mono- and di-methylation. Our previous results demonstrated that mocetinostat decreases LSD1 expression, and the expression of several other lysine demethylases of the jumonji (JmjC) domain class and as a result we observed increases in all types of histone lysine methylation in K562 cells.11 In this study, we found that mocetinostat produced similar increases in all types of histone lysine methylation in MOLM-13 cells. In addition, all compounds increase lysine acetylation at low doses (Figure 3A) but larger doses eliminated this increase with EPZ-5676, while producing even larger increases with mocetinostat (Figure 3B). 4.4 Specific histone H3 modifications Although the pattern of total histone modifications can give some insight into the shared effects of mocetinostat, EPZ-5676, and GSK2879552 (which have different histone modifying enzyme targets),



Page 10 of 31

10 histone modifications elicit different context dependent consequences on gene expression with some modifications increasing, while others decrease transcription. For example, H3K4Me3 increases, while H3K9Me3 decreases transcription, therefore measuring changes in total trimethyl-lysine cannot distinguish the mark from being transcriptionally permissive or repressive.25 Moreover, total histone lysine acetylation and methylation were quantified to observe unanticipated changes in histone modification during treatment with inhibitors of histone modifying enzymes, which could then be investigated more thoroughly with additional experiments. With this in mind, we supplemented our total histone modification data with specific histone H3 modifications using a Histone H3 Modification Multiplex Assay Kit (Abcam) (Figure 3C). Consistent with its total changes in histone modifications, mocetinostat increases lysine methylation and acetylation at a number of residues on histone H3. EPZ5676 produced a generalized decrease in histone H3 modifications with some exceptions including, H3K27Me3 and H3K14. GSK2879552 did not produce the expected increases in H3K4 methylation,26 but rather increased H3K9Me2 and unexpectedly increased H3K79Me1 and 3. Consistent with the total histone acetyl-lysine data, GSK2879552 produced increases in H3K14 and K56 acetylation. Increased H3 acetylation with mocetinostat was expected, however, increases in almost all H3 modifications were also observed. The largest of these were increases in H3K4, H3K9, H3K27 and H3K36 methylation, H3K14 acetylation and H3 serine 28 phosphorylation. Notably, all inhibitors tested appear to produce decreases in H3K79Me2 in MOLM-13 cells. In the context of the pathology of mixed lineage leukemia, decreased H3K79Me2 most likely results from decreased methylation activity by, or expression of DOT1L, as it is the only enzyme known to methylate this residue. In the HOXA9 promoter of cells with the MLL-AF9 fusion, this would result in decreased HOXA9 expression and is consistent with our finding that all compounds tested reduce HOXA9 expression. 4.5 EPZ-5676 reduces LSD1 expression The increases in total histone mono- and dimethyl lysine produced by EPZ-5676 suggest that it may be directly inhibiting or decreasing the expression of LSD1 through an epigenetic mechanism since its


Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


11 effects on total histone modifications had a similar trend as seen with direct inhibition of LSD1 by GSK2879552. To ensure there was no unintentional direct inhibition of LSD1, we tested the activity of mocetinostat, EPZ-5676, and GSK2879552 against DOT1L and LSD1 in vitro and found that EPZ-5676 was not a direct inhibitor of LSD1, but as expected GSK2879552 was (Figure 4). Moreover, we found that GSK2879552 did not directly inhibit DOT1L, but similar to other studies, EPZ-5676 potently inhibited DOT1L.9, 27 Therefore, we used qPCR to determine if EPZ-5676 could reduce LSD1 expression and we found it did so in a time and dose dependent fashion, similar to mocetinostat (Figure 5A). Most interestingly, the maximum reduction in expression of LSD1 by EPZ-5676 resulted from the longest treatment period (144 h), which is a similar time-line to its activity in reducing HOXA9 expression (Figure 5B). Moreover, the magnitude of the IC50 is similar to its reduction in HOXA9 expression (Table 1). These similarities suggest that the reductions in LSD1 and HOXA9 expression we observed may be causally linked. 4.6 EPZ-5676 reduces the expression of a panel of histone modifying enzymes in MOLM-13 cells Establishing that EPZ-5676 decreases many histone H3 modifications, suggested that these changes in histone modifications might be a result of downstream changes in the expression of genes coding for histone modifying enzymes. We therefore measured changes in expression of several histone-modifying enzymes (that are known for their activity as co-activators or co-repressors), in MOLM-13 cells treated with the selected inhibitors (Figure 6). We chose dosing based on each inhibitors concentration and incubation period that resulted in maximum reductions of HOXA9 expression, this being 100 nM, 24 h for mocetinostat, 5 µM, 144 h for EPZ-5676 and 1 µM, 72 h for GSK2879552. Genes from most classes of histone modifying enzymes including lysine methyltransferases and demethylases, lysine acetyltransferases and deacetylases as well as arginine methyltransferases were selected to correlate the changes in expression with the total and sequence specific changes in histone modifications. Treatment with EPZ-5676 resulted in generalized reductions of nearly all genes evaluated, except KAT5 and HDAC5, which remained unchanged in MOLM-13 cells, while some genes appeared to have elevated expression in K562 cells such as HDAC5. A similar pattern was observed with GSK2879552 where there



Page 12 of 31

12 is generally no change or a slight decrease in expression in MOLM-13 cells but a few genes with elevated expression in K562 cells (Figure 6). Reductions in gene expression were generally observed in both cell types with mocetinostat treatment, however increases in KDM4A, HDAC5 and 6, and SIRT2 and 6 were observed in K562 cells. There was a common increase in expression of HDAC5 and decreases in SIRT1 in MOLM-13 and K562 cells with mocetinostat treatment. Consistent with the increases in histone lysine mono- and demethylation (Figure 3), EPZ-5675 and mocetinostat cause reductions in several KDMs but the largest decrease was in LSD1. With respect to mocetinostat, reduction in KDM expression is consistent with our previous results in K562 cells.11 Mocetinostat produces a substantial decrease (83%) in DOT1L expression, which may in part explain the apparent activity of mocetinostat against MOLM-13 cells. In a broad sense the there are many changes in the expression of histone modifying enzymes that are similar upon treatment with EPZ-5676 or mocetinostat. Briefly, lysine acetyltransferase KAT2A, arginine methyltransferases PRMT5, CARM1, lysine methyltransferases EZH2, SET7, SMYD2, DOT1L and lysine demethylase LSD1 all appear to show decreases in expression. In contrast, GSK2879552 shows few changes in the expression of histone modifying enzymes in MOLM-13 cells, however, as with the other compounds there us a notable modest decrease in DOT1L expression. 5. Discussion 5.1 Reductions in HOXA9 expression require different concentrations and dosing periods depending on the epigenetic target Different doses and incubation periods were chosen for each inhibitor based on the doses and times that produced the maximum reduction in HOXA9 expression as measured by dose response curves for decreases in HOXA9 expression in MOLM-13 cells at various incubation times from 24 through 144 h. IC50 values were determined at incubation times that produced the maximum decrease in HOXA9 (Table 1) and these data were used to justify the doses and incubation periods for subsequent experiments. GSK2879552 reached its maximum effect by 72 h and longer incubation periods did not result in greater effects. The decrease in HOXA9 expression induced by EPZ-5676 was ~80% at 72 h, however, there was


Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


13 an additional ~10% decrease in expression up to 144 h, nearly eliminating HOXA9 expression. In order to measure this longer incubation, the cells would need to be passaged, re-seeded to their initial cell number and dosed a second time, which introduced too much error for a small decrease in HOXA9 expression. The reduction in HOXA9 expression produced by mocetinostat plateaued at 24 h and, incubation beyond 48 h at doses higher than 100 nM induced complete cell death in MOLM-13 cells, so examination of the effects beyond 24 h were avoided. 5.2 Common changes in histone modifications and histone modifying gene expression among LSD1, DOT1L and HDAC class I inhibitors. It is difficult to reconcile the diverse inhibitory activities of mocetinostat, EPZ-5676, and GSK2879552 and their shared effect of decreasing HOXA9 expression, however, we did find some common features among all. Each inhibitor either directly inhibits or decreases the expression of LSD1 (Figure 4 and 5). On a global level, this appears to lead to increases in histone lysine mono- and di-methylation (Figure 3A and 3B), and in particular, the repressive mark H3K9Me2 increases (Figure 3C). Although EPZ-5676 directly inhibits DOT1L (Figure 4), EPZ-5676 and all of the other inhibitors we tested also appear to decrease the expression of DOT1L (modestly in the case of GSK2879552) (Figure 6). This likely leads to the observed decrease in H3K79Me2 (Figure 3C) produced by all three inhibitors. This is noteworthy as H3K79Me2 is only catalyzed by DOT1L and previous studies have used the H3K79Me2 mark as a measure of the activity of DOT1L.9 Therefore, specific decreases in H3K79Me2, even with increases of total histone lysine dimethylation (Figure 3A, B), are likely a direct effect of inhibition of, or decreasing expression of DOT1L, which we observed with all compounds. Other studies have suggested that the H3K79Me3 mark is associated with heterochromatin and may be repressive.28, 29 At lower doses, all three inhibitors appear to generally increase total histone acetylation and in particular, all three result in increases of the generally permissive mark H3K14Ac. This effect was expected with the HDAC inhibitor mocetinostat, but it is an off-target effect for EPZ-5676 and GSK2879552, indicating that these inhibitors may alter the expression of HDACs or lysine acetyltransferases. At low doses all inhibitors decrease



Page 14 of 31

14 arginine mono-methylation and asymmetric di-methylation, however, no consistent effect was seen in CARM1 expression, the only enzyme in the gene expression array capable of causing such changes. We have previously shown that mocetinostat increases mono- di- and trimethyl-lysine in K562 cells because mocetinostat reduces the expression of at least seven different lysine demethylases including both FAD dependent and JmjC domain containing types.11 Here we observed similar increases in total histone mono- di- and trimethyl-lysine and a general increase in most specific histone H3 methylation sites (except H379Me2 and H3K27Me2) produced by mocetinostat in MOLM-13 cells (Figure 3). This is likely through its mechanism of decreasing the expression of several lysine demethylases (LSD1, KDM2A and 4A) (Figure 6). On the other hand GSK2879552, and EPZ-5676, increase total histone lysine mono- and di-methylation (Figure 3A and B) but only increase specific histone H3 lysine methylation such as H3K9Me2 and H3K27Me3 (Figure 3C). 5.3 Mocetinostat is a broad-spectrum inducer of apoptosis Mocetinostat is under clinical investigation for a variety of solid tumor cancers in combination with other agents (NCT02805660, NCT02236195, NCT02303262), was withdrawn from a study with azacitidine for acute myeloid leukemia (AML) (NCT00666497) and was part of a study for patients with relapsed or refractory Hodgkin's lymphoma (NCT00358982). Recently the HDAC inhibitors panobinostat, romidepsin and mocetinostat in combination with cytarabine have been shown to augment DNA damaging effects of cytarabine for infant acute lymphoblastic leukemia presenting with MLLrearrangements.30 Our results suggest that mocetinostat alone may be sufficient as a potential treatment for mixed lineage leukemia through its activity of reducing LSD1, DOT1L and HOXA9 expression, and potently reducing MOLM-13 cell viability via apoptosis. In fact, we found that mocetinostat inhibits HOXA9 expression similar to the LSD1 inhibitor GSK2879552 that is currently in dose escalation trials for acute myeloid leukemia (NCT02177812). Although less potent, mocetinostat’s effect of reducing HOXA9 expression is dose dependent and reaches a maximal effect similar to GSK2879552 in MOLM13 cells (Figure 1A). The total reduction in HOXA9 expression in MOLM-13 cells occurred rapidly in 24 h, where GSK2879552 and EPZ-5676, required a minimum of 72 h to observe a substantial reduction in


Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15 HOXA9 expression (Figure 1B). While our results show that mocetinostat, EPZ-5676, and GSK2879552 all decrease HOXA9 expression in an MLL-AF9 translocation cell line, all compounds produce changes in the expression of many genes including those coding for enzymes responsible for adding or removing histone modifications and this leads to changes in histone modifications not associated with the direct enzymatic inhibitory activity of the compounds. With respect to viability, mocetinostat was the most rapid and potent compound we tested, producing apoptosis in both MOLM-13 and K562 cell lines in 24 hours or less. Initial studies with mocetinostat also found it had a broad-spectrum activity inducing apoptosis in colon and lung cancer cell lines.31, 32 More recent studies suggest that mocetinostat induces apoptosis in prostate cancer cells by activation of the microRNA miR-31 and by suppressing expression of the transcription factor E2F6.33 Mocetinostat has also demonstrated some success alone or in combination in clinical trials for relapsed lymphoma and some types of leukemia often through the induction of apoptosis.34-37 Therefore, our study adds to the growing body of evidence that mocetinostat is a broad-spectrum anticancer agent that induces apoptosis. We find for the first time that mocetinostat is a potent inducer of apoptosis in MLL-AF9 cells. 5.4 EPZ-5676 and GSK2879552 reduce MOLM-13 viability through a non-apoptotic mechanism Whereas mocetinostat reduced viability in both MOLM-13 and K562 cells in 24 h by inducing apoptosis, EPZ-5676 and GSK2879552 at an equivalent dose only affected the viability of MOLM-13. Neither EPZ5676, nor GSK2879552 induced caspase activation during their 72 h or 144 h dosing period and their reductions in viability over these times were therefore, via non-apoptotic mechanisms. This was an interesting result as previous groups have shown that knockdown of HOXA9 using shRNA reduces viability in cells with the MLL-AF9 translocation, inducing cell death via apoptosis,10 and that inhibition of DOT1L with EPZ-5676 for 6 days or more in MV4-11 cells, a MLL-AF4 fusion cell line, leads to apoptosis.9 Our results indicated that the inhibition of growth and viability with EPZ-5676 and GSK2879552 in MOLM-13 cells is initially a non-apoptotic mechanism and longer incubation periods may be required to observe apoptosis. In general apoptotic mechanism of induced cell death are more



Page 16 of 31

16 desirable to non-apoptotic because the latter results in inflammation or mediators of inflammation that may produce damage to neighboring healthy cells.38 5.5 Mocetinostat and EPZ-5676 reduce LSD1 expression The finding that both mocetinostat and EPZ-5676 reduce LSD1 expression, suggests that DOT1L may be involved in regulating the expression of LSD1 in MOLM-13 cells and by extension mixed lineage leukemia caused by the MLL-AF9 fusion. EPZ-5676 reduced HOXA9 and LSD1 expression to similar extents over time suggesting the two events were linked. Considering this we evaluated ChIP-seq data for H3K79Me2 at the HOXA9 and LSD1 genes in MLL-AF9 cells with and without treatment with the DOT1L inhibitor EPZ-4777.23 Inhibition of DOT1L resulted in diminished H3K79Me2 in the HOXA9, LSD1, and DOT1L genes (Supplemental Figure S2A-C). When combined with our results that EPZ-5675 reduces the expression of these genes it suggests that DOT1L mediated H3K79Me2, in part, regulates the expression HOXA9, LSD1 and even DOT1L itself (Figures 1, 5 and 6). For mocetinostat, DOT1L and LSD1 expression were simultaneously and substantially reduced in MOLM-13 and K562 cells. In the case of MOLM-13, this resulted in a maximum reduction of ~50% in HOXA9 expression, making it a more effective treatment than GSK2879552. Mocetinostat reduced DOT1L expression by ~83% but the extent to which it reduces HOXA9 expression is not as high as with EPZ-5676. We suspect that the remaining DOT1L activity with mocetinostat treatment is sufficient to sustain some HOXA9 expression. When we knocked down DOT1L expression using siRNA (Supplemental Figure S3) we found that a 70% knockdown sustained for up to 72 hours was insufficient to significantly decrease HOXA9 expression. Therefore, a small amount of DOT1L activity can sustain elevated levels of HOXA9 in MOLM-13 cells. This may explain why EPZ-5676 was the most effective compound we studied at reducing HOXA9 expression because it both directly inhibits and reduces the expression of DOT1L. Alternately, the short duration of treatment with mocetinostat needed to reduce HOXA9 expression may have been insufficient to allow the potential mechanism of reduction of DOT1L expression to be fully realized, as EPZ-5676 requires 144 h for maximum affect. Unfortunately, treatment


Page 17 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


17 of MOLM-13 cells beyond 24h with mocetinostat resulted in reduced cell growth that did not yield sufficient RNA or histones to explore this further. 5.6 Reductions in HOXA9 expression mediated by DOT1L inhibition take multiple cellular generations to appear As with other groups, we found that there was delayed onset of the effects on gene expression with DOT1L inhibitors. This may be due to the apparent lack of an H3K79 demethylase. Although some studies suggest a H3K79 demethylase exists, it has yet to be identified.39 Therefore, in order to effectively reduce H3K79 methylation, cells must undergo nucleosomal turn over and assembly while under persistent DOT1L inhibition to prevent H3K79 methylation on newly assembled chromatin. Treatment with EPZ-5676 must therefore be maintained throughout multiple generations of cells for hypomethylation of H3K79 to occur. Although this may explain the delayed effects observed with EPZ5676, we cannot yet explain why mocetinostat has similar, yet more rapid effects. Previous studies with mocetinostat in colorectal carcinoma cells show rapid cell cycle arrest after only 16 hours.31 However, it is possible that the reduction in HOXA9 expression produced by mocetinostat is mediated by its HDAC inhibition as our previous work has shown that significant increases in histone acetylation are observed in as little as 8 h with mocetinostat.11 5.7 Genome wide H3K79 methylation is generally higher in K562 than MOLM-13: consequences for EPZ-5676 toxicity MOLM-13 cells are particularly sensitive to EPZ-5676 yielding large decreases in the expression of HOXA9, LSD1 and many other genes resulting in potent decreases in cell viability after 144 h treatment. This phenomenon was only observed in MOLM-13 and not in K562 cells, and may be explained by the overall higher levels of H3K79 methylation in K562 versus MOLM-13 cells. This can be observed when comparing H3K79Me2 ChIP-seq data between the two cell lines, with K562 having far higher counts of this mark throughout its genome compared to MLL-AF9 cells (supplemental Figure S2D). (We assume this comparison to be appropriate, as these two individual data sets have been controlled for total input DNA.) This is also in agreement with the finding that MLL-rearranged leukemias have relatively low



Page 18 of 31

18 levels of H3K79Me2 that are reversed upon differentiation.40 Therefore, the lower levels of H3K79Me2 in MOLM-13 cells may make them particularly sensitive to DOT1L inhibitors when compared to K562. Interestingly, the levels of H3K79Me2 within the LSD1 gene in MLL-AF9 cells (Supplemental Figure S2B) are abolished upon treatment with DOT1L inhibitors corresponding with the decreased LSD1 expression we observed (Figure 5). This may not have happened in K562 cells because the local levels of H3K79Me2 in the LSD1 gene are much higher in comparison to those found in MLL-AF9 cells. This may, in part explain the activity of EPZ-5676 in MOLM-13, but not K562 cells. 6. Conclusions Previous studies have shown that inhibition of DOT1L or LSD1 reduce HOXA9 expression and may be useful treatments for mixed lineage leukemia. Here we show that three compounds with different epigenetic targets (mocetinostat (HDAC class I); EPZ-5676, (DOT1L); GSK2879552, (LSD1)), reduce MOLM-13 cell viability by directly inhibiting, or reducing the expression of, HOXA9, DOT1L and LSD1. This results in increases in total histone lysine methylation and acetylation and specifically reduces H3K79Me2 and increase H3K14Ac. Therefore, the efficacy of EPZ-5676 may not be solely a function of its DOT1L enzyme inhibition but may be augmented by its reductions in LSD1 and DOT1L expression, which in turn further reduced HOXA9 expression. GSK2879552 directly inhibits LSD1, and yields a modest reduction in DOT1L expression which further reduces HOXA9 expression. Although not a direct inhibitor of either DOT1L or LSD1, mocetinostat reduces LSD1, DOT1L and HOXA9 expression in MOLM-13 cells and may be another potential treatment for mixed lineage leukemia. As predicted, all three compounds changed the expression of several histone-modifying enzymes which themselves are involved in controlling gene expression. Where this results in reduction of HOXA9 expression these changes in expression are advantageous for the treatment of mixed lineage leukemia. However, where such activities do not eventually decrease HOXA9 expression, they can be viewed as off-target effects. Associated Content Supporting information The following supporting information is available free of charge at the ACS website http://pubs.acs.org


Page 19 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


19 Supplemental file: “Supporting information Lillico JPR” contains supplemental Table 1, Supplemental Figures S1-S5, and supplemental Materials and Methods with additional information on cells lines. Author information *

Corresponding author

Ted M. Lakowski 750 McDermot avenue Winnipeg, Manitoba, Canada, R3E 0T5 204-272-3173 [email protected] Author contributions RL designed and performed experiments, analyzed data and wrote the manuscript; CKL performed experiments and wrote the manuscript; TML designed experiments, analyzed data and wrote the manuscript. Declarations The authors declare no potential conflicts of interest. Acknowledgements The authors acknowledge support from the Dr. Paul H.T. Thorlakson Foundation Fund, The Manitoba Medical Services Foundation [8-2014-03 2014], University of Manitoba Research Grants Program [UM Project # 43887, 2015] and the University Collaborative Research Program (UCRP) [UM Project # 48819, 2018], the Natural Sciences and the Engineering Research Council (NSERC) Canada [RGPIN2015-06543, 2015-2020] (to T.M.L), the Leslie F. Buggey Graduate Scholarship (2014) the Beatrice Faiman Graduate Scholarship and the Research Manitoba Fellowship awarded to R.L. Abbreviations DOT1L, disruptor of telomere silencing 1-like; LSD1, lysine specific demethylase 1; HDAC: histone deacetylase; HOXA9, homeobox protein A9; H3K79, histone 3 lysine 79; H3K4, histone 3 lysine 4; AcK, acetyl-lysine; MeK, monomethyl-lysine; Me2K, dimethyl-lysine; Me3K, trimethyl-lysine; aMe2R, asymmetric dimethyl-arginine; sMe2R, symmetric dimethyl-arginine; MeR, monomethyl-arginine.



Page 20 of 31

20 References

(1) Wong, M.; Polly, P.; Liu, T. The histone methyltransferase DOT1L: regulatory functions and a cancer therapy target. American Journal of Cancer Research. 2015, 5, 2823-2837. (2) McLean, C. M.; Karemaker, I. D.; van Leeuwen, F. The emerging roles of DOT1L in leukemia and normal development. Leukemia. 2014, 28, 2131-2138. (3) Min, J.; Feng, Q.; Li, Z.; Zhang, Y.; Xu, R. M. Structure of the catalytic domain of human DOT1L, a non-SET domain nucleosomal histone methyltransferase. Cell. 2003, 112, 711-723. (4) Nguyen, A. T.; Zhang, Y. The diverse functions of Dot1 and H3K79 methylation. Genes & development. 2011, 25, 1345-1358. (5) Morera, L.; Lübbert, M.; Jung, M. Targeting histone methyltransferases and demethylases in clinical trials for cancer therapy. Clinical Epigenetics. 2016, 8, 57. (6) Krivtsov, A. V.; Armstrong, S. A. MLL translocations, histone modifications and leukaemia stem-cell development. Nat Rev Cancer. 2007, 7, 823-833. (7) Slany, R. K. The molecular biology of mixed lineage leukemia. Haematologica. 2009, 94, 984-993. (8) Kuntimaddi, A.; Achille, N. J.; Thorpe, J.; Lokken, A. A.; Singh, R.; Hemenway, C. S.; Adli, M.; Zeleznik-Le, N. J.; Bushweller, J. H. Degree of Recruitment of DOT1L to MLL-AF9 Defines Level of H3K79 Di- and Tri-methylation on Target Genes and Transformation Potential. Cell reports. 2015, 11, 808-820. (9) Daigle, S. R.; Olhava, E. J.; Therkelsen, C. A.; Basavapathruni, A.; Jin, L.; Boriack-Sjodin, P. A.; Allain, C. J.; Klaus, C. R.; Raimondi, A.; Scott, M. P.; Waters, N. J.; Chesworth, R.; Moyer, M. P.; Copeland, R. A.; Richon, V. M.; Pollock, R. M. Potent inhibition of DOT1L as treatment of MLL-fusion leukemia. Blood. 2013, 122, 1017-1025. (10) Faber, J.; Krivtsov, A. V.; Stubbs, M. C.; Wright, R.; Davis, T. N.; van den Heuvel-Eibrink, M.; Zwaan, C. M.; Kung, A. L.; Armstrong, S. A. HOXA9 is required for survival in human MLL-rearranged acute leukemias. Blood. 2009, 113, 2375-2385. (11) Lillico, R.; Sobral, M. G.; Stesco, N.; Lakowski, T. M. HDAC inhibitors induce global changes in histone lysine and arginine methylation and alter expression of lysine demethylases. J Proteomics. 2016, 133, 125-133. (12) Lillico, R.; Stesco, N.; Khorshid Amhad, T.; Cortes, C.; Namaka, M. P.; Lakowski, T. M. Inhibitors of enzymes catalyzing modifications to histone lysine residues: structure, function and activity. Future medicinal chemistry. 2016, 8, 879-897. (13) Amente, S.; Lania, L.; Majello, B. The histone LSD1 demethylase in stemness and cancer transcription programs. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms. 2013, 1829, 981-986. (14) Wang, J.; Scully, K.; Zhu, X.; Cai, L.; Zhang, J.; Prefontaine, G. G.; Krones, A.; Ohgi, K. A.; Zhu, P.; Garcia-Bassets, I.; Liu, F.; Taylor, H.; Lozach, J.; Jayes, F. L.; Korach, K. S.; Glass, C. K.; Fu, X.-D.; Rosenfeld, M. G. Opposing LSD1 complexes function in developmental gene activation and repression programmes. Nature. 2007, 446, 882-887. (15) Kerenyi, M. A.; Shao, Z.; Hsu, Y.-J.; Guo, G.; Luc, S.; O'Brien, K.; Fujiwara, Y.; Peng, C.; Nguyen, M.; Orkin, S. H. Histone demethylase Lsd1 represses hematopoietic stem and progenitor cell signatures during blood cell maturation. eLife. 2013, 2, e00633. (16) Harris, W. J.; Huang, X.; Lynch, J. T.; Spencer, G. J.; Hitchin, J. R.; Li, Y.; Ciceri, F.; Blaser, J. G.; Greystoke, B. F.; Jordan, A. M.; Miller, C. J.; Ogilvie, D. J.; Somervaille, T. C. P. The Histone Demethylase KDM1A Sustains the Oncogenic Potential of MLL-AF9 Leukemia Stem Cells. Cancer cell. 2012, 21, 473-487.


Page 21 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


21 (17) Feng, Z.; Yao, Y.; Zhou, C.; Chen, F.; Wu, F.; Wei, L.; Liu, W.; Dong, S.; Redell, M.; Mo, Q.; Song, Y. Pharmacological inhibition of LSD1 for the treatment of MLL-rearranged leukemia. Journal of Hematology & Oncology. 2016, 9, 24. (18) Kim, W.; Choi, M.; Kim, J.-E. The histone methyltransferase Dot1/DOT1L as a critical regulator of the cell cycle. Cell Cycle. 2014, 13, 726-738. (19) Nguyen, A. T.; Taranova, O.; He, J.; Zhang, Y. DOT1L, the H3K79 methyltransferase, is required for MLL-AF9-mediated leukemogenesis. Blood. 2011, 117, 6912-6922. (20) Scoumanne, A.; Chen, X. The Lysine-specific Demethylase 1 Is Required for Cell Proliferation in Both p53-dependent and -independent Manners. Journal of Biological Chemistry. 2007, 282, 15471-15475. (21) Foster, C. T.; Dovey, O. M.; Lezina, L.; Luo, J. L.; Gant, T. W.; Barlev, N.; Bradley, A.; Cowley, S. M. Lysine-Specific Demethylase 1 Regulates the Embryonic Transcriptome and CoREST Stability. Molecular and Cellular Biology. 2010, 30, 4851-4863. (22) Beck, B.; Chen, Y. F.; Dere, W.; Devanarayan, V.; Eastwood, B. J.; Farmen, M. W.; Iturria, S. J.; Iversen, P. W.; Kahl, S. D.; Moore, R. A.; Sawyer, B. D.; Weidner, J. Assay Operations for SAR Support; (In) G. S. Sittampalam, N. P. Coussens and K. Brimacombe: Assay Guidance Manual, Eli Lilly & Company and the National Center for Advancing Translational Sciences; : Bethesda (MD), 2012; pp 1161-1176. (23) Riedel, S. S.; Haladyna, J. N.; Bezzant, M.; Stevens, B.; Pollyea, D. A.; Sinha, A. U.; Armstrong, S. A.; Wei, Q.; Pollock, R. M.; Daigle, S. R.; Jordan, C. T.; Ernst, P.; Neff, T.; Bernt, K. M. MLL1 and DOT1L cooperate with meningioma-1 to induce acute myeloid leukemia. The Journal of clinical investigation. 2016, 126, 1438-1450. (24) Lawrence, H. J.; Rozenfeld, S.; Cruz, C.; Matsukuma, K.; Kwong, A.; Komuves, L.; Buchberg, A. M.; Largman, C. Frequent co-expression of the HOXA9 and MEIS1 homeobox genes in human myeloid leukemias. Leukemia. 1999, 13, 1993-1999. (25) Hyun, K.; Jeon, J.; Park, K.; Kim, J. Writing, erasing and reading histone lysine methylations. Experimental &Amp; Molecular Medicine. 2017, 49, e324. (26) Mohammad, H. P.; Smitheman, K. N.; Kamat, C. D.; Soong, D.; Federowicz, K. E.; Van Aller, G. S.; Schneck, J. L.; Carson, J. D.; Liu, Y.; Butticello, M.; Bonnette, W. G.; Gorman, S. A.; Degenhardt, Y.; Bai, Y.; McCabe, M. T.; Pappalardi, M. B.; Kasparec, J.; Tian, X.; McNulty, K. C.; Rouse, M.; McDevitt, P.; Ho, T.; Crouthamel, M.; Hart, T. K.; Concha, N. O.; McHugh, C. F.; Miller, W. H.; Dhanak, D.; Tummino, P. J.; Carpenter, C. L.; Johnson, N. W.; Hann, C. L.; Kruger, R. G. A DNA Hypomethylation Signature Predicts Antitumor Activity of LSD1 Inhibitors in SCLC. Cancer cell. 2015, 28, 57-69. (27) Daigle, S. R.; Olhava, E. J.; Therkelsen, C. A.; Majer, C. R.; Sneeringer, C. J.; Song, J.; Johnston, L. D.; Scott, M. P.; Smith, J. J.; Xiao, Y.; Jin, L.; Kuntz, K. W.; Chesworth, R.; Moyer, M. P.; Bernt, K. M.; Tseng, J. C.; Kung, A. L.; Armstrong, S. A.; Copeland, R. A.; Richon, V. M.; Pollock, R. M. Selective killing of mixed lineage leukemia cells by a potent small-molecule DOT1L inhibitor. Cancer cell. 2011, 20, 53-65. (28) Barski, A.; Cuddapah, S.; Cui, K.; Roh, T. Y.; Schones, D. E.; Wang, Z.; Wei, G.; Chepelev, I.; Zhao, K. High-resolution profiling of histone methylations in the human genome. Cell. 2007, 129, 823-837. (29) Chantalat, S.; Depaux, A.; Héry, P.; Barral, S.; Thuret, J.-Y.; Dimitrov, S.; Gérard, M. Histone H3 trimethylation at lysine 36 is associated with constitutive and facultative heterochromatin. Genome Research. 2011, 21, 1426-1437.



Page 22 of 31

22 (30) Cruickshank, M. N.; Ford, J.; Cheung, L. C.; Heng, J.; Singh, S.; Wells, J.; Failes, T. W.; Arndt, G. M.; Smithers, N.; Prinjha, R. K.; Anderson, D.; Carter, K. W.; Gout, A. M.; Lassmann, T.; O'Reilly, J.; Cole, C. H.; Kotecha, R. S.; Kees, U. R. Systematic chemical and molecular profiling of MLL-rearranged infant acute lymphoblastic leukemia reveals efficacy of romidepsin. Leukemia. 2017, 31, 40. (31) Fournel, M.; Bonfils, C.; Hou, Y.; Yan, P. T.; Trachy-Bourget, M. C.; Kalita, A.; Liu, J.; Lu, A. H.; Zhou, N. Z.; Robert, M. F.; Gillespie, J.; Wang, J. J.; Ste-Croix, H.; Rahil, J.; Lefebvre, S.; Moradei, O.; Delorme, D.; Macleod, A. R.; Besterman, J. M.; Li, Z. MGCD0103, a novel isotype-selective histone deacetylase inhibitor, has broad spectrum antitumor activity in vitro and in vivo. Molecular cancer therapeutics. 2008, 7, 759-768. (32) Zhou, N.; Moradei, O.; Raeppel, S.; Leit, S.; Frechette, S.; Gaudette, F.; Paquin, I.; Bernstein, N.; Bouchain, G.; Vaisburg, A.; Jin, Z.; Gillespie, J.; Wang, J.; Fournel, M.; Yan, P. T.; Trachy-Bourget, M. C.; Kalita, A.; Lu, A.; Rahil, J.; MacLeod, A. R.; Li, Z.; Besterman, J. M.; Delorme, D. Discovery of N-(2-aminophenyl)-4-[(4-pyridin-3-ylpyrimidin-2ylamino)methyl]benzamide (MGCD0103), an orally active histone deacetylase inhibitor. J Med Chem. 2008, 51, 4072-4075. (33) Zhang, Q.; Sun, M.; Zhou, S.; Guo, B. Class I HDAC inhibitor mocetinostat induces apoptosis by activation of miR-31 expression and suppression of E2F6. Cell Death Discovery. 2016, 2, 16036. (34) Younes, A.; Oki, Y.; Bociek, R. G.; Kuruvilla, J.; Fanale, M.; Neelapu, S.; Copeland, A.; Buglio, D.; Galal, A.; Besterman, J.; Li, Z.; Drouin, M.; Patterson, T.; Ward, M. R.; Paulus, J. K.; Ji, Y.; Medeiros, L. J.; Martell, R. E. Mocetinostat for relapsed classical Hodgkin's lymphoma: an open-label, single-arm, phase 2 trial. The Lancet. Oncology. 2011, 12, 1222-1228. (35) Garcia-Manero, G.; Assouline, S.; Cortes, J.; Estrov, Z.; Kantarjian, H.; Yang, H.; Newsome, W. M.; Miller, W. H., Jr.; Rousseau, C.; Kalita, A.; Bonfils, C.; Dubay, M.; Patterson, T. A.; Li, Z.; Besterman, J. M.; Reid, G.; Laille, E.; Martell, R. E.; Minden, M. Phase 1 study of the oral isotype specific histone deacetylase inhibitor MGCD0103 in leukemia. Blood. 2008, 112, 981-989. (36) Batlevi, C. L.; Crump, M.; Andreadis, C.; Rizzieri, D.; Assouline, S. E.; Fox, S.; van der Jagt, R. H. C.; Copeland, A.; Potvin, D.; Chao, R.; Younes, A. A phase 2 study of mocetinostat, a histone deacetylase inhibitor, in relapsed or refractory lymphoma. British journal of haematology. 2017, 178, 434-441. (37) Blum, K. A.; Advani, A.; Fernandez, L.; Van Der Jagt, R.; Brandwein, J.; Kambhampati, S.; Kassis, J.; Davis, M.; Bonfils, C.; Dubay, M.; Dumouchel, J.; Drouin, M.; Lucas, D. M.; Martell, R. E.; Byrd, J. C. Phase II study of the histone deacetylase inhibitor MGCD0103 in patients with previously treated chronic lymphocytic leukaemia. British journal of haematology. 2009, 147, 507-514. (38) Haanen, C.; Vermes, I. Apoptosis and inflammation. Mediators of Inflammation. 1995, 4, 5-15. (39) Farooq, Z.; Banday, S.; Pandita, T. K.; Altaf, M. The many faces of histone H3K79 methylation. Mutation research. Reviews in mutation research. 2016, 768, 46-52. (40) Wong, S. H. K.; Goode, D. L.; Iwasaki, M.; Wei, M. C.; Kuo, H.-P.; Zhu, L.; Schneidawind, D.; Duque-Afonso, J.; Weng, Z.; Cleary, M. L. The H3K4-methyl epigenome regulates leukemia stem cell oncogenic potential. Cancer cell. 2015, 28, 198-209.


Page 23 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


23

Tables Table 1. IC50 values for reduction in HOXA9 and LSD1 expression, and cell viability upon treatment with EPZ-5676, mocetinostat and GSK2879552. HOXA9 expression IC50 (nM)

*

max/min (% expression) LSD1 expression IC50 (nM) max/min (% expression) MOLM-13 viability IC50 (nM) max/min (% viability) K562 viability IC50 (nM)

EPZ-5676

mocetinostat

GSK2879552

150±10

105±53

10±3

101±12/22±3

89±5/51±2

90±3/58±1

172±40

127±18

N/A

91±12/57±5

91±2/38±4

N/A

196±88

109±17

5±1

98±7/28±7

106±14/19±1

124±14/69±8

N/A

409±25

N/A

104±3/11±1

110±2/95±6

max/min (% viability) 100±2/96±5 *measurements are means of 3 measurements ± SD Figure Captions

Figure 1. Inhibitors of histone modifying enzymes decrease HOXA9 expression depending on treatment time and dose in MOLM-13 cells. Displayed are dose response curves measured by qPCR (N=3) showing decreasing HOXA9 expression with increasing concentration of the LSD1 inhibitor GSK2879552 after 72 h, the HDAC inhibitor mocetinostat after 24h, and the DOT1L inhibitor EPZ-5676 after 72 h (A). Time points were chosen to elicit the maximum effect over the minimum time for one dose of each inhibitor. The maximum reduction of HOXA9 gene expression with a single dose of 1 µM GSK2879552 or 5 µM EPZ-5676 required longer incubation times (72 h) compared to 100 nM mocetinostat, which the maximum reduction on HOXA9 expression occurred at 24 h of treatment (B). Values represent the mean of 3 measurements and error bars represent SD.

Figure 2. EPZ-5676, mocetinostat and GSK2879552 reduce cell growth by different mechanisms. Changes in cell viability are displayed for a 10 µM dose of GSK2879552 for 72 h, mocetinostat for 24 h and EPZ-5676 for 144 h in the MLL-AF9 cell line MOLM-13 and K562 cells (A). Percent changes were measured as a ratio with respect to a no treatment control for both cell lines. Apoptosis was measured by caspase activation with a DMSO control, 1 µM GSK2879552 for 72 h, EPZ-5676 for 144 h or mocetinostat for 24 h. Only mocetinostat appears to induce apoptosis (B). The percentage of viable cells with increasing concentrations of GSK2879552 for 72 h, EPZ-5676 for 144 h and mocetinostat for 24 h in MOLM-13 (C), and K562 cells (D). Viability and apoptosis were measured using the Apo-Tox Glo cell assay. Values represent the mean of 3 measurements and error bars represent SD.

Figure 3. The changes in histone modifications in cells treated with GSK2879552, EPZ-5676 and mocetinostat. Shown are the changes in total histone lysine and arginine modifications observed with 100nM treatment of mocetinostat for 24h, 1 µM of GSK2879552 for 72 h and a 3 µM of EPZ-5676 for 144 h in MOLM-13 cells measured using the LC-MS/MS assay described in materials and methods (A). Similar changes in total histone modifications were observed with 1 µM of mocetinostat for 24, 3 µM of GSK2879552 for 72 h and a 5 µM dose of EPZ-5676 for 144 h (B). Bars represent means with SD of 3 measurements. A heat-map of changes to specific histone H3 modifications measured using the Histone



Page 24 of 31

24 H3 Modifications Multiplex Assay Kit (abcam) in MOLM-13 cells using a 5 µM EPZ-5676 100 nM mocetinostat and 1 µM GSK2879552 at the times above (C). Each cell of the heat-map is the mean of 4 measurements. Supplemental Figure 4 shows the same data with the corresponding heat-map of the error expressed a CV% calculated as CV%=(SD/mean)100%.

Figure 4. EPZ-5676 and GSK2879552 inhibit DOT1L and LSD1, respectively in vitro. 10 µM of the inhibitors were used in either the LSD1 inhibitor screening assay (Cayman) or the LC-MS/MS assay adapted to in vitro measurement of DOT1L methylation activity as outlined in Materials and Methods. Mocetinostat did not inhibit LSD1 or DOT1L in vitro (data not shown). Values are means, and error bars SD for 3 measurements.

Figure 5. EPZ-5676 and mocetinostat reduce LSD1 gene expression in MOLM-13 cells. MOLM-13 cells were treated with increasing doses of EPZ-5676 for 72 h or mocetinostat for 24 h (A) and LSD1 expression was measured by qPCR. The reductions in LSD1 expression induced by EPZ-5676 was measured at varying time-points with maximum reductions of both LSD1 ~80% and HOXA9 expression >90% observed after 144 h (B). Maximum reduction in LSD1 expression induced by mocetinostat appear at 24 h as we have previously shown. 11 Values represent the mean of 3 measurements and error bars represent SD.

Figure 6. A gene expression heat map of selected histone modifying enzymes in K562 and MOLM-13 cells treated with mocetinostat, EPZ-5676, or GSK2879552. Treatments of 100 nM mocetinostat for 24h, 5 µM EPZ-5676 for 144h and 1 µM GSK2879552 for 72h were evaluated in MOLM-13 and the control K562 cells for changes in expression of histone modifying enzymes, measured by qPCR. Each cell of the heat-map is the mean of 3 measurements. Supplemental Figure 5 shows the same data with the corresponding heat-map of the error expressed a CV% calculated as CV%=(SD/mean)100%.


Page 25 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


25 For Table of Contents graphic Only



Figure 1. Inhibitors of histone modifying enzymes decrease HOXA9 expression depending on treatment time and dose in MOLM-13 cells. Displayed are dose response curves measured by qPCR (N=3) showing decreasing HOXA9 expression with increasing concentration of the LSD1 inhibitor GSK2879552 after 72 h, the HDAC inhibitor mocetinostat after 24h, and the DOT1L inhibitor EPZ-5676 after 72 h (A). Time points were chosen to elicit the maximum effect over the minimum time for one dose of each inhibitor. The maximum reduction of HOXA9 gene expression with a single dose of 1 µM GSK2879552 or 5 µM EPZ-5676 required longer incubation times (72 h) compared to 100 nM mocetinostat, which the maximum reduction on HOXA9 expression occurred at 24 h of treatment (B). Values represent the mean of 3 measurements and error bars represent SD. 139x241mm (300 x 300 DPI)


Page 26 of 31

Page 27 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. EPZ-5676, mocetinostat and GSK2879552 reduce cell growth by different mechanisms. Changes in cell viability are displayed for a 10 µM dose of GSK2879552 for 72 h, mocetinostat for 24 h and EPZ-5676 for 144 h in the MLL-AF9 cell line MOLM-13 and K562 cells (A). Percent changes were measured as a ratio with respect to a no treatment control for both cell lines. Apoptosis was measured by caspase activation with a DMSO control, 1 µM GSK2879552 for 72 h, EPZ-5676 for 144 h or mocetinostat for 24 h. Only mocetinostat appears to induce apoptosis (B). The percentage of viable cells with increasing concentrations of GSK2879552 for 72 h, EPZ-5676 for 144 h and mocetinostat for 24 h in MOLM-13 (C), and K562 cells (D). Viability and apoptosis were measured using the Apo-Tox Glo cell assay. Values represent the mean of 3 measurements and error bars represent SD. 117x81mm (300 x 300 DPI)



Figure 3. The changes in histone modifications in cells treated with GSK2879552, EPZ-5676 and mocetinostat. Shown are the changes in total histone lysine and arginine modifications observed with 100nM treatment of mocetinostat for 24h, 1 µM of GSK2879552 for 72 h and a 3 µM of EPZ-5676 for 144 h in MOLM-13 cells measured using the LC-MS/MS assay described in materials and methods (A). Similar changes in total histone modifications were observed with 1 µM of mocetinostat for 24, 3 µM of GSK2879552 for 72 h and a 5 µM dose of EPZ-5676 for 144 h (B). Bars represent means with SD of 3 measurements. A heat-map of changes to specific histone H3 modifications measured using the Histone H3 Modifications Multiplex Assay Kit (abcam) in MOLM-13 cells using a 5 µM EPZ-5676 100 nM mocetinostat and 1 µM GSK2879552 at the times above (C). Each cell of the heat-map is the mean of 4 measurements. Supplemental Figure 4 shows the same data with the corresponding heat-map of the error expressed a CV% calculated as CV%=(SD/mean)100%. 123x90mm (300 x 300 DPI)


Page 28 of 31

Page 29 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. EPZ-5676 and GSK2879552 inhibit DOT1L and LSD1, respectively in vitro. 10 µM of the inhibitors were used in either the LSD1 inhibitor screening assay (Cayman) or the LC-MS/MS assay adapted to in vitro measurement of DOT1L methylation activity as outlined in Materials and Methods. Mocetinostat did not inhibit LSD1 or DOT1L in vitro (data not shown). Values are means, and error bars SD for 3 measurements. 61x50mm (300 x 300 DPI)



Figure 5. EPZ-5676 and mocetinostat reduce LSD1 gene expression in MOLM-13 cells. MOLM-13 cells were treated with increasing doses of EPZ-5676 for 72 h or mocetinostat for 24 h (A) and LSD1 expression was measured by qPCR. The reductions in LSD1 expression induced by EPZ-5676 was measured at varying timepoints with maximum reductions of both LSD1 ~80% and HOXA9 expression >90% observed after 144 h (B). Maximum reduction in LSD1 expression induced by mocetinostat appear at 24 h as we have previously shown. 11 Values represent the mean of 3 measurements and error bars represent SD. 140x245mm (300 x 300 DPI)


Page 30 of 31

Page 31 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. A gene expression heat map of selected histone modifying enzymes in K562 and MOLM-13 cells treated with mocetinostat, EPZ-5676, or GSK2879552. Treatments of 100 nM mocetinostat for 24h, 5 µM EPZ-5676 for 144h and 1 µM GSK2879552 for 72h were evaluated in MOLM-13 and the control K562 cells for changes in expression of histone modifying enzymes, measured by qPCR. Each cell of the heat-map is the mean of 3 measurements. Supplemental Figure 5 shows the same data with the corresponding heatmap of the error expressed a CV% calculated as CV%=(SD/mean)100%. 147x174mm (300 x 300 DPI)


Selective DOT1L, LSD1, and HDAC class I inhibitors reduce HOXA9

Recommend Documents